Learned motor sequences underlie most of human communication, yet remarkably little is known about how the nervous system learns to control the complex muscle actions involved. Our long term goal is to describe the neural circuit mechanisms underlying the acquisition of learned motor behaviors. The zebra finch, a songbird, provides a unique system in which to pursue this goal, as it acquires its song in much the same way that we learn many of our motor skills, including speech. Aims: Our proposal aims to describe how the motor program for song develops by recording from neurons in a motor cortex analogue structure (nucleus RA) in the freely behaving, juvenile zebra finch throughout song learning (Aim 1). Widely thought to be the site of vocal learning, RA receives convergent input from a higher order motor area, HVC, and from a basal ganglia circuit. The proposal will examine the respective roles of these two inputs in shaping the motor command in RA during learning (Aim 2). We examine how auditory feedback-based performance evaluation, a crucial ingredient for both song learning and song maintenance, influences the development of the motor program (Aim 3). Lastly, we assess the extent to which the learning induced changes in the RA motor program are driven by changes in HVC, its premotor input (Aim 4). Methods: The proposal will examine these issues with a combination of powerful methods: custom- made motorized microdrives will allow the recording of single neurons in RA in the singing, juvenile bird, and a chronically implanted reverse microdialysis device will make possible the fast and reversible inactivation of the basal ganglia circuit. Finally, we will use mathematical models that incorporate our observations into a biophysically plausible model of how the song circuit learns and functions. PUBLIC HEALTH RELEVANCE Our experiments aim to describe how the motor program underlying a complex motor behavior evolves, and the logic by which the motor circuits underlying it are organized with respect to learning. The homologies and analogies between the neural circuits generating vocalizations in songbirds and humans are many, thus our findings will also speak to the question of how the motor program underlying speech and other learned motor behaviors may be acquired. Understanding the neural correlates of complex motor learning will allow us to pinpoint how the process may fail, thus addressing the possible causes of various motor disorders and disabilities.